by Grania
The Oatmeal weighs in on Google self-driving cars, which as far as I am concerned can’t arrive soon enough given that humans and their goddamned “intuition” manage to kill 1.2 million and injure another 50 million humans every year.
It’s still a work in progress, but it is so very much better than this:
Read his whole post here.
We’ll have self-flying airplanes first, because 1) it’s a lot simpler, and 2) we’re pretty much there already.
We already have the pilots trained to feed the dog – and the dog trained to bite the pilot if he tries to touch the controls.
Sub
They’re still trying to perfect automatic soap dispensers. (Or, in case they think they’ve perfected them, they would be mistaken. Not that this was anything society was crying out for, either.)
Also, come back when barcode scanners work 100% of the time.
I think it’s a given that self-driving cars would occasionally make mistakes which would lead to death or injury. The best argument that can be made, though, is that they would make a lot fewer mistakes than humans would.
However, I think humans tend to be more forgiving of human mistakes than computer mistakes.
Add to that the unlikely influence of alcohol being a factor with the robotic device.
Bender would disagree.
Actually, an alcohol-fuelled robot isn’t that insane an idea – given an appropriate fuel cell. The burning belches would be for homage/ effect only.
If anything my understanding is the driverless car is too cautious – if that’s a word – eg, the obstacle avoidance
function can’t distinguish between a cinder block and a discarded newspaper.
That’s because sensor technology as currently applied to driverless cars is woefully inadequate. Same as the navigation issues because GPS units (not GPS itself note) rely on missing and inaccurate databases. There are probably solutions, if not now then soon, but will it be cost effective?
Long distance haulage completely automated and driverless within ten years? I’m skeptical.
But what will the railroad workers do then?
Exactly.
A robot car doesn’t need to win every Formula 1 race.
A robot car just has to be a slightly better driver than the average human is when that human is, say, getting a passing score on a driving test.
The robot is never going to get sleepy, never going to get drunk, never put on makeup, never check out that cute girl on the sidewalk, never be fuming at the stupid thing the boss just demanded, never have to tell the kids that we’re not there yet, and on and on and on. All those times when humans are at their deadliest, the robot is still going to be just as safe as the human is at the human’s best.
What few people realize is that robot vehicles are going to completely revolutionize society.
For starters, long-haul trucking is going to be completely automated a decade from now. A robot trucker can run 24/7 — a single truck can do in one day what it’d take three drivers to legally do. The trucks only have to stop load, unload, and refuel…and the latter can be done while they’re being loaded. And they can have their “home bases” be a quarter mile from the freeway, and so not have to bother with all the messiness of city traffic. The first package delivery firm to go robot will be able to cut prices in half and simultaneously double profits, and thereby eat the lunch of every other firm that doesn’t immediately follow.
Yes, unions will stand in the way. But not for long. If necessary, one of the big firms will start a new small “independent” company that’s pure robot, and then merge when the robot company is big enough. Or it could be a truly independent upstart, or Google, or whomever. Point is, truckers are the new buggy whips.
Passenger service will rapidly follow.
People will still own cars for quite a while, but they’ll soon be primarily for recreation and entertainment and status and the like. It’ll be cheaper, faster, and more convenient, especially for commuters, to punch in a request for a robot car on the smartphone. Look for payment plans like with cellphones…free unlimited mileage for a low payment of $x/month, and so on.
The remaining technical hurdles are trivial compared to the regulatory hurdles, and those will crumble as soon as some rich company decides to buy enough politicians. The marketing firms will take care of social resistance.
…and, oh-by-the-way, it’s not just driving that robots are about to take over, but all sorts of other jobs that nobody realizes are prime for automation. Medicine, for example…the robot doesn’t have to be the best general practitioner on the planet; it only has to be better than average. It really isn’t that much longer before all kinds of jobs are simply automated out of existence, and there’s nothing else left for those people to do for a job.
And then how’re we all going to pay the bills? And who’s going to collect the pay the robots get that used to go to humans?
Very interesting times ahead, indeed.
b&
The first phone calls I made went through the operator. The first banks I used had no ATMs. I no longer use a travel agent to plan a trip or buy my airline tickets (and I make dinner reservations on different continents from home).
Auto-drive automobile? Just gives me more time to read whyeveloutiontrue on my iPad.
The future sounds terrific. What are we waiting for?
Flying cars to get us there?
Uh, buggy-whip before the horses ass?
Hoverboards.
“but all sorts of other jobs that nobody realizes are prime for automation.”
Well, all jobs can arguably be done by robots of sufficient sophistication. This is likely to produce social problems in the future. Some people have made the argument that when technology makes some jobs obsolete, it always creates more opportunity than it destroys. I think this argument is naive. As a nation, we’ve lost millions of high paying manufacturing jobs and replaced them with low paying service jobs. What happens when those jobs disappear?
Exactly.
In the beginning, it was true…but we’re way past the beginning.
I myself have automated a number of accounting-type jobs out of existence. What, exactly, is the new opportunity for an accountant who’s been automated out of a job? What’s all that education prepared her for other than accounting, and how does she pay off her student loans for that now-useless education?
And I fully expect my own job to be automated out of existence at some point, and with minimal warning, when everything I do gets outsourced to ADP or some other “Cloud” provider. I’m ready for that in the sense that I know it’s coming and I’ve got other plans lined up to keep Baihu fed…but I also know full well that I’ll likely never again be paid as well as I am today.
One way or another, we’re going to have to figure out what to do with all these people who used to fully contribute to society but who, through no fault of their own, have had their best or even only means of self-sufficiency replaced by a robot.
…and it’s going to move very rapidly up the food chain, too…middle managers are on the chopping block, too, once all the jobs they used to manage get automated out of existence….
b&
Assuming we’re on the same page here with regard to humans being nothing more than extraordinarily sophisticated robots driven by deterministic principles, no matter the pace of advancement, so long as it is in a forward direction, we’re all ultimately on the chopping block.
I’ve fooled around with some meta-template programming in C++. It’s complicated as hell and frankly new graduates prefer web technologies with more simplistic APIs. Nonetheless, someone has to maintain “the automated part” of the process. While programming myself out of existence is several orders of magnitude over what we have now (call it meta-meta-meta-programming if you want to), it is bound to happen. Once quantum computing is perfected and we have parallel processing with more processing power than the brain, we’re all done. Sam Harris has talked about this before, and he’s right. Within the next century, under the current scheme, we could have a quadrillionaire while also having 40% unemployment. We’re several decades away now, but this is what the future looks like if we don’t come up with a new economic scheme.
First, government employment figures are horribly skewed by the fact that they don’t count as unemployed the long-term unemployed, only those who qualify for unemployment insurance. The numbers of people actually employed…mean rather scary high actual unemployment figures.
That aside, we’ll never actually make it to quadrillionaires with massive unemployment, because, if nothing else, a significant fraction of the unemployed will riot rather than starve.
Not that the types of social change that come from food riots are something to look forward to. There’s a good chance it’ll be even more authoritarian, at least at first. But it does mean there’re some limits to extrapolation.
https://xkcd.com/605/
b&
Well, so long as Xkcd says so, I stand corrected. In all seriousness, we’re not that far removed from a trillionaire. The idea that someone comes up with an idea more pervasive than FB or Microsoft isn’t far fetched. Hell, with appropriate funding we could put the retail and fast food service industry out of business. There’s a huge chunk of my 40% scenario right there. Do you think that an invention that results in the obselesence of fast food workers couldn’t result in a trillionaire (or at least the equivalent in corporate profit)? Now imagine the next big breakthrough. If it is out of the U.S. with our low tax rates, I can see the profits going another 1000x higher. Buffet’s wealth alone could fund a million middle class families for a lifetime, and he didn’t even come up with a revolutionary product.
All else equal, sure.
But what makes you think all those now-unemployed fast food workers are just going to peacefully starve to death while this trillionaire snorts coke off the tits of a dozen hookers?
And, without jobs, who’s going to have money to buy robot-prepared fast food?
b&
Sure, it wouldn’t be pretty. But history shows that stark class divides between rich and poor never are. We already have Apple heading towards being a trillion dollar company and much of their profit is due to the obscenely cheap offshore labor they use. With companies growing at this rate, an individual worth a trillion dollars within the next couple of decades isn’t out of the question. I was hyperbolizing in my initial statement about a quadrillionaire, though an economic disaster resulting in severe inflation could create such a scenario.
I think we already saw a microcosm of what you allude to with the OWS protests, and you’re right, a situation that puts some few tens of millions of retail and restaurant workers out of work would result in more than just a few hippies in a park with no direction as to what the hell they’re trying to accomplish.
I would like to think we learned our history lesson with the Great Depression, but the Republicans, especially the current crop of Tea Party Members, either seem to have forgotten what widespread poverty looks like, or they don’t care. The fact that nearly a century after Teddy Roosevelt went on his trust busting campaign, we once again had companies that were “too big to fail” doesn’t bode well for understanding the need for social safety nets.
Frankly, I can see our current crop of politicians vigorously arguing that the trillionaire founders of Company X who put all these people out of work deserve their money and shouldn’t be taxed more. And, the base would go right along with it as usual, voting against their own interests. We would likely sort things out ultimately, but not without a lot of turmoil as we adjust to a new normal.
Woops, sent too early. Meant to caveat that with third world countries. Anyhow, at 50 billion, that’s 2 billion a year in returns in a conservative investment, even in the US, that’s 10000 families covered for 40 years at median pay.
Pet rant coming.
I have always (since I got my driving license – at age 24) been of the opinion that people who pass a driving test should be issued with their driving license for a fixed period (10 years as a point ; open to discussion). Then (this is the heretical bit) they lose their license and must go through that year’s mandatory new-driver instruction and pass the test current in that year. Lather, rinse and repeat.
Vehicles do change ; driving rules do change ; expected driver behaviours do change. Therefore driver qualifications should also change, regardless of whether the driver in question has currently got a license or not.
I tried one of the UK Govt’s “propose some legislation” things (I forget what it was called) and I think it acquired 8 votes. Not a convincing movement for change. But I still think it’s necessary.
Doing this with robot cars is almost trivial. “Management of change” is an issue for all life-threatening industries, but it’s also not exactly rocket science. It’s a significant component of rocket science, but not the whole thing.
Makes sense to me. In aviation the FAA has moved toward mandatory retraining – the biannual flight review and a lot of non-mandatory training opportunities. Not exactly like taking the whole training package over again, but a worthwhile mechanism for change control. For me the biggest change has been the use of the glass panel display. The steam gauges I leaned on are simpler, but the digital displays and computer based navigation provide much more information. Information overload is an issue.
“biannual flight review ”
But the mandatory training has failed to teach people that it’s no longer called a “Biennial Flight Review”, but simply “Flight Review”. This has been the case for over 15 years, but even flight instructors still refer to BFRs.
Yep, know those feelings.
I’ve always said the same thing here in the United States. Especially after having spent a portion of my life in Florida, where retirees flock to sit in the sun and aimlessly meander across all lanes of the road while driving 15 mph under the speed limit, I see no reason why driver’s tests should not be mandatory every few years. The problem is that these senior citizens make up a significant voting block and apply enough pressure so that politicians don’t even want to touch the issue; the implicit message, of course, is that they want to keep driving even if they are incapable of passing the test. I don’t doubt there’s some percentage of people in all age groups who think this way, especially when we insist on releasing obnoxious chart topping songs entitled, Jesus Take the Wheel.
Advanced Warning: Watching the above video may cause bouts of nausea, anger, desire to have a stiff drink, or all of the above.
Jesus take the Wheel? Something like this happened here in NY a few years ago. The woman was a bit manic and left home with 4 or 5 kids in the car. I guess Jesus was busy because they ended up wrapped around a tree. Mom and most of her kids died. Reality is sometimes harsh.
Was this the incident on the Taconic Parkway? I recall the mother was allegedly high and I seem to recall her having a head-on collision and all the kids in the car getting killed (as well as I think everyone in the other vehicle).
I’ll skip the video.
If the crinklies are going to block it, fine : introduce it for the pre-melanoma phase of humanity. Then let the courts force the “grandfathering age up to something more reasonable. If you’re over 178, then you don’t need to re-sit your test.
It’s politics. I’ll get the nigre lye and scrub my brain clean now.
Several obstacles to that:
1) People consider driving a right, not a privilege.
2) It would vastly increase the cost of administration.
3) Wouldn’t improve safety at all.
(1) They’re wrong. tough.
(2) Not that much really. It’s welded into the system. No ifs, no buts, no maybes. You are issued with a license with an expiry date. It expires. You’re not a legal driver any more. End of debate. No appeal procedure. The only complication would be – does your legally-required insurance expire at the same time as the license.
(3)Elaborate. It’ll scare thousands of drivers off the road. I can’t see how that would be a bad thing. If they’re replaced by professional drivers (with appropriate additional training, and enforcement ; I gather that enforcement is incredibly lax in the States, but that’s not necessarily a problem with the rules, but with their enforcement.) or mechanical drivers, all the better.
I’m with you on your rant.
Hell, the last driving test I myself took was as a teenager in a different state. And the expiration date on my license is a couple decades away…and all I’ll have to do to renew it is get a new photo and maybe take an eye exam.
b&
Well, we now have to renew photos on IIRC a 10-year basis. (Which, of course, involves a charge.) But no mention that I’ve heard of that you’ll be required to pass an eye test.
There are a variety of disease which doctors are required to report to the DVLA if they’re diagnosed AND the doctor has reason to believe that the patient has a driving license. Which is about 35% odds for 17-20 and peaks at 80/905 for females/ males in their 40s. But very often the doctor doesn’t make a report for various reasons including believing the patient’s assurance that they’ll surrender their license themselves.
Once the design is tested and proven, people will grow accustomed to it. Hell, they already have. How many vehicle owners complain that they’re worried that the CVT can’t shift gears as well as they can or that the ABS Pumps the brakes rather than waiting for you to manually do it when you enter a skid? Same goes for more recent additions such as detection systems for when you drift out of your lane.
I suspect after a few years drifting out of your lane won’t be an issue. The road way will provide guidance in the form of something like an wire under the pavement or some kind of laser beam thingy. We may even give up windows in favor of a virtual view of wherever we want to pretend we are.
I wonder how strong EMF (electromagnetic field) these cars, particularly their radars, generate, having in mind the issue of potential overexposure to such RF fields.
http://www.who.int/peh-emf/publications/facts/fs226/en/
Eeesh. Do they use radar? I assumed it’s all lasers and imagery analysis.
“Google’s driverless car tech uses an array of detection technologies including sonar devices, stereo cameras, lasers, and radar”
http://www.extremetech.com/extreme/189486-how-googles-self-driving-cars-detect-and-avoid-obstacles
I would think they use LIDAR (only IR light probably). In any event, the RF powers would be negligible for a car’s immediate survellience systems.
The link you provide doesn’t support your implication that there are actual cases of overexposure to RF fields. The guide lines for keeping below particular beam powers are well known and are not constraints on design or implementation of appropriate range systems. No sane engineer would design a system that uses hundreds or thousands of times the power necessary to achieve the task desired.
To quote your citation on the scariest allegation :
The WHO summary mentions several potential problems caused by RF fields:
“Non-thermal effects: Exposure to RF levels too low to involve heating, (i.e., very low SARs), has been reported by several groups to alter calcium ion mobility, which is responsible for transmitting information in tissue cells. However, these effects are not sufficiently established to provide a basis for restricting human exposure.
Pulsed RF fields: Exposure to very intense pulsed RF fields, similar to those used by radar systems, has been reported to suppress the startle response and evoke body movements in conscious mice. In addition, people with normal hearing have perceived pulse RF fields with frequencies between about 200 MHz and 6.5 GHz. This is called the microwave hearing effect. The sound has been variously described as a buzzing, clicking, hissing or popping sound, depending on the RF pulsing characteristics. Prolonged or repeated exposure may be stressful and should be avoided where possible.”
When it comes to RF and cancer link, it says that “further studies are necessary.”
I just wonder what levels of EMF these self-driving cars generate, and if they were ever to go big, could millions of radars on the streets pose any health risk.
WHO is probably not your best source on this. There is literally practically nothing not on their ‘suspected possible carcinogen’ list. (and EVERYTHING is ‘further research is needed’)
Credible research has failed to produce any real issues (hundreds of millions of cell phones over decades have not be accompanied by a cancer surge).
Unfortunately, the wingnut sides of the health and environmental movements will undoubtedly mount an ‘anti-vax’ style PR campaign. Apparently they’re more worried about low RF exposure than getting hit by a car.
“Credible research has failed to produce any real issues (hundreds of millions of cell phones over decades have not be accompanied by a cancer surge).”
The issue is not so clear-cut as you believe, and the industry-sponsored research is seriously lacking in credibility
“In 2009, the Journal of Clinical Oncology published the findings of a team of seven scientists, who reviewed 23 epidemiological studies on the link between cell phone use and cancer. The team concluded that
‘although as a whole the data varied, among the 10 higher quality studies, we found a harmful association between phone use and tumor risk. The lower quality studies, which failed to meet scientific best practices, were primarily industry funded. The 13 studies that investigated cell phone use for 10 or more years found a significant harmful association with tumor risk, especially brain tumors, giving us ample reason for concern about long-term use.’
Martin Blank PhD – Overpowered: The Dangers of Electromagnetic Radiation (EMF) and What You Can Do about It
First point :
I take that as “damned close to too small to measure”.
Second point :
So you don’t do that then. It may be relevant in this context that I’m sitting almost a metre from my hard hat – with 104dB rated ear defenders attached. I put 95dB rated ear plugs in before clamping the ear defenders on.
There are good reasons that high-power RF sources are on one side of perimeter fences and the general public are on the other side of those fences. It’s to keep people away from the high power areas, until the beam leaves the landscape or drops below acceptable power levels.
To run at those power levels, your car would need to be towing a fairly large (several tonne) generator pack for each high-power radar set. That’s another extremely good reason for keeping the power levels used as low as possible. (I’m now trying to remember if the wife insisted on having air conditioning in the new car – I’d not bother, but stuck in a traffic jam she considers it important. OTOH, she also insisted on a sun roof (in Scotland!!) and I cannot remember if I persuaded her to forego the power waste of the AC. Your high power radars would render that sort of debate moot.
Specifically Electro-Motive Force? – that would be a few microvolts per metre, falling off on a 1/r-squared law. Probably lower than the EMF coming from a mobile phone, which still have no proven deleterious side effects. Lots of allegations, no generally accepted evidence. I lump it with the antivaxxers, but without the wider health harms that the antivaxxers have.
“Specifically Electro-Motive Force?”
No, specifically the EMF generated by these cars on-board radars.
“Probably lower than the EMF coming from a mobile phone, which still have no proven deleterious side effects.”
Here’s another interesting quote from the Overpowered book:
“A 2007 review of 16 studies on this subject found that the studies all demonstrated increased risk of brain tumors known as glioma and acoustic neuroma among cell phone users. Compiling the data across studies for 10 years and greater, the risk of developing an ipsilateral glioma tumor (a tumor on the same side of the head as the cell phone is used) was elevated 240%. The incidence of ipsilateral tumor reinforces the connection between radiation and cancer, as these tumors develop in the precise area where exposure to cell phone radiation has been most intense.
[…]
In 2009, Dr. Lennart Hardell reported that children who begin using mobile phones at ages younger 20 have a 520% elevated risk of developing glioma”
All sorts of things can be extracted from data if you try hard enough (especially if you have an agenda).
The simpler question is: with hundreds of millions of phones in daily use, where is the epidemic? (If you argue that there has been too little time, then how did the ‘evidence’ happen to come up in the smaller data sample?)
Smaller data sample? These are findings coming from the metaanalysis of *multiple* epidemiological studies.
[Yawn]
I’ll just step over the bodies of people fried in the gutters by their mobile phones. What? An absence of gutters blocked with bodies.
At very high levels of emitted power there are good reasons to be careful of anything – be it gamma rays, alpha particles or radio waves. Or, for that matter sound waves or any sort of flammable material. But at lower intensities, they literally fade into the background. Our ancestors have lived with (small) RF fluxes (and other radiation fluxes) and we’re well acclimatised to living with them for long enough to produce the next generation. Much beyond that, unless it is literally causing people to die in the streets (I’m thinking of that mercury-in-fish poisoning case -“Minimata disease” or something similar), then it’s probably not that much of a problem.
OTOH, it’s taken a couple of hours to write this up because I do have something interesting to deal with, so I’ll have to return to the question at the end of my shift.
Ca ion mobility. What kind of physics are they talking about? RF that is competitive with chemical, diffusion processes? Very unlikely and almost certain unmeasurable.
Think of the reverse problem. Has anyone set up an experiment to manipulate, stimulate, inhibit cell activity with RF? Here’s one paper that’s orthogonal to the wishes of those who think that RF can cause harm:
http://www.nature.com/bjc/journal/v106/n2/full/bjc2011523a.html
In the below presentation Dr Devra Davis talks about a study showing that RF can cause DNA damage.
You do realize grocery store automated doors and motion sensor lights are “radars”, used to detect people at about the same range(distance)?
Most motion sensor lights are passive systems (they don’t generate any energy) which detect IR energy. As for radar-based motion detectors, are you suggesting that since they are used on automatic doors, there’s no concern about introducing millions of new radars on the streets and thus increasing people’s exposure to EMF?
No. I am suggesting that people have been living with these devices for decades and haven’t been worried that the door was going to kill them.
The thing is the exposure to RF in this case is limited to what, 30 seconds or 2 minutes? And with many cars in the traffic equipped with radars this exposure could increase significantly.
BTW, I will have to measure SAR / MPE at my place one of these days, because my computer with built-in Wi-Fi detects 20 other wireless networks, of which 9 have maximum strength.
I think they will have it on the Tesla before long…at least driver assist. I would really like one if someone would pay for it.
I’m in the computer biz, and I remain deeply skeptical of this. Google cars operated on road go by extremely detailed internal maps and GPS in a very restricted list of settings. That is not going to change soon.
Driver assist is where the real action will be. People tend to over-stimate what computers can do (particularly where actual intelligence is needed) and under-estimate the complexity of real world driving. In a crowded intersection, a computer cannot tell the difference between a pedestrian who’s made eye contact with a driver and one who is looking the other way.
Emergency braking, blind spot warning, etc are areas for real improvements in safety without requiring actual AI (which the more I think about it, the farther off it seems– Watson has a huge information base to draw on, but is, by any human standard other than dataset size, pretty stupid).
There are some interesting driver assist technologies coming up though in the off road area as well, Land Rover is working on sonar that can ‘see’ underwater obstacles (great for water crossings), and supposedly it will plot an ideal path through a rock strewn passage maintaining optimal clearance and stability. From my own experience over the years, that has been a real challenge to learn (I don’t get stuck much any more).
Right. The equipment in my plane (Garmin panel) gives a visual and audible ( synthetic voice )warning when another aircraft is close enough to pose a risk. This is the kind of thing cars will do before they become fully automatic.
I think any serious notion of having self driving cars has to require a major infrastructure upgrade as well. I see it as useful in major urban areas like here in New York. If we removing driver reaction time from the scenario, gridlock massively improve. But I don’t see it as remotely plausible without an infrastructure such as networked roadways to support it; i.e. the cars can have some built in intelligence but there needs to be a command center(s) similar to that of air traffic control for this to work. The roadways would have to be cordoned off from bicycles and pedestrians, think tunnels or overpasses at every intersection in order to simplify calculations to the vehicles on the network rather than trying to take into account thousands of degrees of freedom that are the pedestrians and their erratic movements.
In our current political climate, this is pure fantasy. It’s hard enough to get funding to even modernize our current rail systems and maintain our bridges. Look no further than the current East Side Access project for an example of the debacle we make out of a major upgrade to the system. Now multiply that by several orders of magnitude to rework Manhattan to support a networked system of autonomous vehicles, and we’re talking being centuries away.
We have reliable technology to upgrade mass transit and greatly reduce the need for personal vehicles and thus reducing the risk for driver error. This is a realistic goal and we can’t even get political backing for that, and when we do it is so poorly planned that the projects run years over the estimate. It would be fantastic if the majority of the population, I’m thinking anyone with 50 miles of a major city was within say 2-3 miles of a train station. Combine this with better biking and pedestrian paths and we could cut our reliance on cars significantly. Our existing roads would be more than enough to support the reduced traffic and reduce our reliance on oil at the same time, but this all takes funding and a hell of a lot more resources that we are currently allocating to infrastructure upgrades (which is none for the majority of the United States).
No need at all. There’re already all sorts of ad-hoc cooperative networking protocols that just require a radio transceiver (such as the one you already carry in your pocket) and little else. Cars will relay all sorts of things to each other. A bunch of cars going to the same destination will “link up” to form a sort of a train, the ones in back tailgating inches off the bumper of the one in front, and the one in front responsible for telling the others how to maneuver. Cars a mile down the road will broadcast a signal about some sort of hazard, and all the ones behind will immediately begin an easy slowdown and start moving to the proper lane — well before any human would even know anything is amiss. Similarly, cars will alert each other about non-robot vehicles behaving erratically so they can give it a wide berth.
And, again, all without any sort of central control.
b&
Overall, this is a win win for bike commuters.
Yeah, there’ll be something else for the pathological bike-haters to hate on.
More importantly than John’s (valid) point, it’ll be fewer pathological bike haters on the roads.
Unless … I’m just remembering the Futurama Santa and transposing it to a world of robot cars and human cyclists. [shudder]
Robot cyclists? If anthropoid robots take off, then there will be times that they need to move from point A to point B. The engineering efficiencies of bicycles over legs for (reasonably) smooth-paved areas still apply. Therefore, if there are anthropoid robots, there will be bicycles for them.
In my neck of the woods, the biggest wins for bike commuters would be no hills and no snow.
Another thought…the insurance companies are going to go nuts.
On the one hand, they’ll be thrilled. They’ll think they can charge basically the same rates to insure a robot car as a non-robot car, maybe with some small discount, and almost never have to pay out.
On the other hand, the cars will mostly be owned and operated by taxi-equivalent companies that’ll self-insure, and suddenly the market for car insurance will evaporate overnight.
“Disruptive” is definitely the key word, here.
b&
Fair points.
I fail to see the need to bemoan the death of the vehicle insurance industry. Where’s my shovel?
Oh, I’m not worried about them in particular. But they do wield a fair amount of political clout, so there could well be some fireworks to enjoy.
b&
In the spirit of the season, let’s throw another dog on the fire, skewer a few creationists (you know they enjoy it really) and set them to sizzle, and the try to figure out if the helicopter’s will be flying on going-home day. Not even 50% odds on today’s weather forecast.
…and, even if they don’t, Sithrak will!
b&
No, not the dogs:-((
Waily waily! Grocers asotrophry!
I’d want one that I could take over from when desired. Say, have it drive me to a wild area, then let me drive, to go down whatever two-track looks interesting at the time, hop out at the drop of a hat, crawl as slow as necessary for whatever it is I’m watching, stop dead whenever I see a good photo op, etc. I don’t see how I could possibly communicate some of those needs quickly enough to a robot.
In these cases my wife and I take turns on who has to watch the ruts in the road and who gets to use the car as a blind(hide) or look for the best photo ops.
What I gather is that they will kick control back to the erstwhile driver if some communication problem is encountered. Great – the e.d. who has been snoozing is suddenly supposed to snap to attention, process everything at once, and make decisions.
No thanks.
Because I’m SO with-it when I just wake up…
Good point!
One more thing no one’s brought up; cars are quite vulnerable to hacking as it is.
That’s only going to increase.
The car can always stop, pull to the side of the road, and make you a cup of coffee.
Better check that the beans are topped up when you plug in the charger. There’s sure to be an app for that.
As someone who is rapidly approaching geezerhood, I really hope self-driving cars become a reality. Really liberating for old folks.
And, of course, I already deeply appreciate the Google search engine when one of those annoying senior memory holes open up. Shortens the torture and embarrassment of not being able to remember the name of something you’ve known since childhood.
Yes, but can they be programmed to give The Finger to other self-driving cars?
These cars will certainly be better than human drivers at using their turn signal. Here in Miami, drivers NEVER use their turn signal — unless the signal was on when they bought the car, in which case they never turn it OFF.
There are going to be a lot of unemployed truckers.
Dear God. Any chance of keeping the damn things out of the fast lane?
I like the Tesla and Mercedes Benz approach: keep adding automated features to traditional-driver cars to take over every aspect of actually driving the vehicle, slowly transitioning the human to be more navigator than operator. Eventually the technology will mature and people will become comfortable enough that true driverless cars will become possible.
You’re far more likely to die on rural roads (which are often narrow, curvy, and have high speed limits) than city streets. If we can get this technology into $20,000 economy cars that wouldl save a ton of lives.
http://www-nrd.nhtsa.dot.gov/Pubs/812101.pdf
Still, props to Google for working on this from the opposite direction. This will be huge for people who can’t drive at all, if and when it ever becomes street-legal. My mom gave up driving years ago and would adore one of these.
I was witness to what may have been the first accident between two autonomous cars. It happened at the third DARPA Grand Challenge when the two vehicles involved approached a “Y” intersection. Both came to a stop, then both moved forward into a very minor fender-bender. I could have happened to two human drivers, except one of them would have waved the other forward.